Early Progress in Rate Modeling
Since beginning your journey towards higher sophistication on the rate modeling side, what are some of the early wins or signs of progress that stood out to you?
For James River, one milestone stands above the rest: leaving fully manual modeling behind. When Randy arrived, modeling work was spread across Python scripts, Jupyter notebooks, and a collection of files with no consistent structure. Collaboration was difficult, and tracking model versions or updates required manual coordination.
Today, the picture looks entirely different. With a centralized modeling platform, the team can now:
- Store all models and data in a single, unified environment
- Manage champion and challenger models more efficiently
- Automate model outputs — eliminating reliance on spreadsheets and manual IT handoffs
“If we change the model, hit a few buttons, what comes out is what our IT department needs.”
Randy Compton, AVP of Data Science & Analytics
This shift has removed operational bottlenecks, reduced errors, and enabled a far more scalable modeling process. Model updates that once required multiple Excel tabs and lengthy coordination now happen with a few clicks. This is accelerating James River’s journey toward a fully automated, data-driven pricing framework.
Measuring ROI in a Full-Scale Pricing Transformation
What does ROI look like when you’re looking at a full-scale pricing transformation?
For Randy Compton, AVP Data Science & Analytics at James River Insurance, the return on investment comes down to two core objectives: speed and precision.
Manual workflows still slow response times, which is a critical factor in James River’s business. The ability to quickly evaluate a submission and identify high-quality risks can create significant competitive advantage. Automation plays a key role: freeing underwriters to prioritize policies with the greatest potential impact.
At the same time, James River is advancing toward more accurate, risk-aligned pricing. Moving past ISO-based approaches allows the company to build technical pricing that better reflects true exposure. The next evolution is experience-based rating, where seemingly similar policies are priced differently based on claims history and risk profile.
The outcome is clear: faster decisions, sharper pricing, and a more profitable book of business.
Key Takeaways for Insurers Embarking on a Pricing Transformation
What is one key takeaway you would share with other insurers embarking on a pricing transformation?
Randy emphasizes one principle above all: a pricing transformation is not an actuarial project, it’s an enterprise project. Pricing touches every corner of the organization, and success requires collaboration across underwriting, actuarial, and operations.
He also shared a core philosophy that guides his work:
“Every policy can be a profitable policy.”
In his view, unprofitable policies are often the result of mispricing, not inevitability. By finding the right price for each individual policy ( supported by data, analytics, and cross-functional input) insurers can create a pricing strategy that is both fair to customers and financially sustainable.
